Abstract
Purpose :
Ophthalmology clinics with minimal statistical analytics capabilities may struggle to implement a quality monitoring system to evaluate patient outcomes. We performed a quality improvement study by designing a quality monitoring system for a low-resource ophthalmology clinic that would use small samples to estimate outcomes for larger patient populations.
Methods :
We evaluated the proportion of primary open-angle glaucoma (POAG) patients who were treated “successfully” according to American Academy of Ophthalmology Preferred Practice Patterns (PPPs). We analyzed 100 patients seen in the clinic over 3 months in 2019 as the input for our Bayesian analysis. We also created a standardized note template for use by clinicians in the electronic medical record (EMR) for POAG patient visits to facilitate monitoring of treatment successes without requiring clinical expertise. We evaluated adherence to clinician template use in POAG patient notes on the weekly day of glaucoma patient visits over 9 weeks in 2020.
Results :
Using Bayesian analysis based on our initial data, we created tables that allow for quality monitoring of future 3 month intervals using parameters from smaller samples. Using a small sample of 30 patients, we were able to determine that there was a 100% probability that the clinic as a whole was not achieving the pre-defined target percentage of 80% of POAG patients successfully treated, as defined by the PPPs. Regarding adherence to use of the note template for POAG patients, the median proportion of patients for which the template was used on a single clinic day was 27%, with a maximum of 53% and a minimum of 11%.
Conclusions :
Based on the input parameters of the number of patients in the sample, the number of treatment successes in that sample, and the target proportion of successfully treated patients, a small sample Bayesian analysis can be used for quality monitoring of patient outcomes for larger patient populations in low-resource clinical settings. Future work includes creation of an open-access database of tables derived from Bayesian analysis for use as a reference by other low-resource clinics in quality monitoring efforts.
This is a 2021 ARVO Annual Meeting abstract.